linux/Documentation/lockup-watchdogs.txt
<<
>>
Prefs
   1===============================================================
   2Softlockup detector and hardlockup detector (aka nmi_watchdog)
   3===============================================================
   4
   5The Linux kernel can act as a watchdog to detect both soft and hard
   6lockups.
   7
   8A 'softlockup' is defined as a bug that causes the kernel to loop in
   9kernel mode for more than 20 seconds (see "Implementation" below for
  10details), without giving other tasks a chance to run. The current
  11stack trace is displayed upon detection and, by default, the system
  12will stay locked up. Alternatively, the kernel can be configured to
  13panic; a sysctl, "kernel.softlockup_panic", a kernel parameter,
  14"softlockup_panic" (see "Documentation/kernel-parameters.txt" for
  15details), and a compile option, "BOOTPARAM_SOFTLOCKUP_PANIC", are
  16provided for this.
  17
  18A 'hardlockup' is defined as a bug that causes the CPU to loop in
  19kernel mode for more than 10 seconds (see "Implementation" below for
  20details), without letting other interrupts have a chance to run.
  21Similarly to the softlockup case, the current stack trace is displayed
  22upon detection and the system will stay locked up unless the default
  23behavior is changed, which can be done through a sysctl,
  24'hardlockup_panic', a compile time knob, "BOOTPARAM_HARDLOCKUP_PANIC",
  25and a kernel parameter, "nmi_watchdog"
  26(see "Documentation/kernel-parameters.txt" for details).
  27
  28The panic option can be used in combination with panic_timeout (this
  29timeout is set through the confusingly named "kernel.panic" sysctl),
  30to cause the system to reboot automatically after a specified amount
  31of time.
  32
  33=== Implementation ===
  34
  35The soft and hard lockup detectors are built on top of the hrtimer and
  36perf subsystems, respectively. A direct consequence of this is that,
  37in principle, they should work in any architecture where these
  38subsystems are present.
  39
  40A periodic hrtimer runs to generate interrupts and kick the watchdog
  41task. An NMI perf event is generated every "watchdog_thresh"
  42(compile-time initialized to 10 and configurable through sysctl of the
  43same name) seconds to check for hardlockups. If any CPU in the system
  44does not receive any hrtimer interrupt during that time the
  45'hardlockup detector' (the handler for the NMI perf event) will
  46generate a kernel warning or call panic, depending on the
  47configuration.
  48
  49The watchdog task is a high priority kernel thread that updates a
  50timestamp every time it is scheduled. If that timestamp is not updated
  51for 2*watchdog_thresh seconds (the softlockup threshold) the
  52'softlockup detector' (coded inside the hrtimer callback function)
  53will dump useful debug information to the system log, after which it
  54will call panic if it was instructed to do so or resume execution of
  55other kernel code.
  56
  57The period of the hrtimer is 2*watchdog_thresh/5, which means it has
  58two or three chances to generate an interrupt before the hardlockup
  59detector kicks in.
  60
  61As explained above, a kernel knob is provided that allows
  62administrators to configure the period of the hrtimer and the perf
  63event. The right value for a particular environment is a trade-off
  64between fast response to lockups and detection overhead.
  65
  66By default, the watchdog runs on all online cores.  However, on a
  67kernel configured with NO_HZ_FULL, by default the watchdog runs only
  68on the housekeeping cores, not the cores specified in the "nohz_full"
  69boot argument.  If we allowed the watchdog to run by default on
  70the "nohz_full" cores, we would have to run timer ticks to activate
  71the scheduler, which would prevent the "nohz_full" functionality
  72from protecting the user code on those cores from the kernel.
  73Of course, disabling it by default on the nohz_full cores means that
  74when those cores do enter the kernel, by default we will not be
  75able to detect if they lock up.  However, allowing the watchdog
  76to continue to run on the housekeeping (non-tickless) cores means
  77that we will continue to detect lockups properly on those cores.
  78
  79In either case, the set of cores excluded from running the watchdog
  80may be adjusted via the kernel.watchdog_cpumask sysctl.  For
  81nohz_full cores, this may be useful for debugging a case where the
  82kernel seems to be hanging on the nohz_full cores.
  83